Facial Expression Recognition Software Developed
Applying the facial expression recognition algorithm, the developed prototype is capable of processing a sequence of frontal images of moving faces and recognizing the person’s facial expression. The software can be applied to video sequences in realistic situations and can identify the facial expression of a person seated in front of a computer screen. Although still only a prototype, the software is capable of working on a desktop computer or even on a laptop.
Flexibility and adaptability
The system analyses the face of a person sitting in front of a camera connected to a computer running the prototype. The system analyses the person’s face (up to 30 images per second) through several boxes, each “attached” to or focusing on part of the user’s face. These boxes monitor the user’s facial movements until they manage to determine what the facial expression is by comparison with the expressions captured from different people (333 sequences) from the Cohn-Kanade database.
The system’s success rate on the Cohn-Kanade database is 89%. It can work under adverse conditions where ambient lighting, frontal facial movements or camera displacements produce major changes in facial appearance.
This software has a range of applications: advanced human-computer interfaces, improved relations with the e-commerce consumers, and metaverse avatars with an unprecedented capability to relate to the person they represent.
This software can enrich advanced human-computer interfaces because it would enable the construction of avatars that really do simulate a person’s facial expression. This is a really exciting prospect for sectors like the video games industry.
Electronic commerce could also benefit from this technology. During the e-commerce buying process, the computer would be able to identify potential buyers’ gestures, determine whether or not they intend to make a purchase and even gauge how satisfied they are with a product or service by helping to reduce the ambiguities of spoken or written language.
Applied to metaverses like Second Life, this software would also enable the avatars representing system users to act out the feelings of the user captured through facial expressions.
Although there are some facial analysis products on the market, none specifically target the analysis of user facial expressions. Visit the Computational Perception and Robotics Research Group’s website for videos illustrating the algorithm in operation.
Additionally, while most similar systems developed by other researchers focus on just part of expression recognition, the developed prototype does the whole job: 1) locates and monitors the face in the image using an algorithm that works despite changes of illumination or user movement, and 2) classifies the user’s facial expression. Finally, it also incorporates an original algorithm that calculates the likely evolution of the analysed user’s facial expressions.
Springer London, ISSN 1433-7541 (Print) 1433-755X (Online) DOI10.1007/s10044-007-0084-8
José M. Buenaposada, Enrique Muñoz and Luis Baumela
Abstract We introduce a system that processes a sequence of images of a front-facing human face and recognises a set of facial expressions. We use an efficient appearance-based face tracker to locate the face in the image sequence and estimate the deformation of its non-rigid components. The tracker works in real time. It is robust to strong illumination changes and factors out changes in appearance caused by illumination from changes due to face deformation. We adopt a model-based approach for facial expression recognition. In our model, an image of a face is represented by a point in a deformation space. The variability of the classes of images associated with facial expressions is represented by a set of samples which model a low-dimensional manifold in the space of deformations. We introduce a probabilistic procedure based on a nearest-neighbour approach to combine the information provided by the incoming image sequence with the prior information stored in the expression manifold to compute a posterior probability associated with a facial expression. In the experiments conducted we show that this system is able to work in an unconstrained environment with strong changes in illumination and face location. It achieves an 89% recognition rate in a set of 333 sequences from the Cohn–Kanade database.